Hong-Mei Chen, R. Kazman, Serge Haziyev, Valentyn Kropov, Dmitri Chtchourov
Big data as a Service (BDaaS) provides a viable strategy for organizations to implement scalable, tailorable big data infrastructure and applications built on this infrastructure. New trends in the BDaaS market are moving toward an open world model -- what we call the Neo-Metropolis model -- for developing BDaaS platforms. The key to the success of such large-scale technology-agnostic platforms, we posit, is an architectural strategy revolving around microservices and DevOps. This article presents the results of an action research with a Neo-Metropolis BDaaS vendor and illustrates how architectural support for DevOps is critical in achieving desired system qualities and enabling platform success. This research contributes to illuminate best practices of DevOps, and to validate and augment a set of DevOps tactics previously developed, while adding and recategorizing new instances of well-established architectural tactics.
{"title":"Architectural Support for DevOps in a Neo-Metropolis BDaaS Platform","authors":"Hong-Mei Chen, R. Kazman, Serge Haziyev, Valentyn Kropov, Dmitri Chtchourov","doi":"10.1109/SRDSW.2015.14","DOIUrl":"https://doi.org/10.1109/SRDSW.2015.14","url":null,"abstract":"Big data as a Service (BDaaS) provides a viable strategy for organizations to implement scalable, tailorable big data infrastructure and applications built on this infrastructure. New trends in the BDaaS market are moving toward an open world model -- what we call the Neo-Metropolis model -- for developing BDaaS platforms. The key to the success of such large-scale technology-agnostic platforms, we posit, is an architectural strategy revolving around microservices and DevOps. This article presents the results of an action research with a Neo-Metropolis BDaaS vendor and illustrates how architectural support for DevOps is critical in achieving desired system qualities and enabling platform success. This research contributes to illuminate best practices of DevOps, and to validate and augment a set of DevOps tactics previously developed, while adding and recategorizing new instances of well-established architectural tactics.","PeriodicalId":415692,"journal":{"name":"2015 IEEE 34th Symposium on Reliable Distributed Systems Workshop (SRDSW)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122743482","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Manel Abdellatif, C. Talhi, A. Hamou-Lhadj, M. Dagenais
Malware detection on mobile phones involves analysing and matching large amount of data streams against a set of known malware signatures. Unfortunately, as the number of threats grows continuously, the number of malware signatures grows proportionally. This is time consuming and leads to expensive computation costs, especially for mobile devices where memory, power and computation capabilities are limited. As the security threat level is getting worse, parallel computation capabilities for mobile phones is getting better with the evolution of mobile graphical processing units (GPUs). In this paper, we discuss how we can benefit from the evolving parallel processing capabilities of mobile devices in order to accelerate malware detection on Android mobile phones. We have designed and implemented a parallel host-based anti-malware for mobile devices that exploits the computation capabilities of mobile GPUs. A series of computation and memory optimization techniques are proposed to increase the detection throughput. The results suggest that mobile graphic cards can be used effectively to accelerate malware detection for mobile phones.
{"title":"On the Use of Mobile GPU for Accelerating Malware Detection Using Trace Analysis","authors":"Manel Abdellatif, C. Talhi, A. Hamou-Lhadj, M. Dagenais","doi":"10.1109/SRDSW.2015.18","DOIUrl":"https://doi.org/10.1109/SRDSW.2015.18","url":null,"abstract":"Malware detection on mobile phones involves analysing and matching large amount of data streams against a set of known malware signatures. Unfortunately, as the number of threats grows continuously, the number of malware signatures grows proportionally. This is time consuming and leads to expensive computation costs, especially for mobile devices where memory, power and computation capabilities are limited. As the security threat level is getting worse, parallel computation capabilities for mobile phones is getting better with the evolution of mobile graphical processing units (GPUs). In this paper, we discuss how we can benefit from the evolving parallel processing capabilities of mobile devices in order to accelerate malware detection on Android mobile phones. We have designed and implemented a parallel host-based anti-malware for mobile devices that exploits the computation capabilities of mobile GPUs. A series of computation and memory optimization techniques are proposed to increase the detection throughput. The results suggest that mobile graphic cards can be used effectively to accelerate malware detection for mobile phones.","PeriodicalId":415692,"journal":{"name":"2015 IEEE 34th Symposium on Reliable Distributed Systems Workshop (SRDSW)","volume":"131 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121739807","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Chenyang Qu, D. Ulybyshev, B. Bhargava, R. Ranchal, L. Lilien
Data exchange between vehicles and base stations may contain information on traffic accidents, traffic jams, road constructions etc. Risks to data privacy in vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) systems need be minimized. We present a policybased solution that provides controlled and privacypreserving dissemination of video data in V2V and V2I. This policy-based solution relies on Active Bundle (AB), which incorporates policy enforcement mechanism and policies that describe access to video data. The usage of ABs ensures privacy of actors during disclosure of captured video in untrusted environment. Face recognition algorithm is used to identify sensitive data in videos and it is used in policies. We use four algorithms to process images that are captured by a vehicle's camera.
{"title":"Secure Dissemination of Video Data in Vehicle-to-Vehicle Systems","authors":"Chenyang Qu, D. Ulybyshev, B. Bhargava, R. Ranchal, L. Lilien","doi":"10.1109/SRDSW.2015.22","DOIUrl":"https://doi.org/10.1109/SRDSW.2015.22","url":null,"abstract":"Data exchange between vehicles and base stations may contain information on traffic accidents, traffic jams, road constructions etc. Risks to data privacy in vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) systems need be minimized. We present a policybased solution that provides controlled and privacypreserving dissemination of video data in V2V and V2I. This policy-based solution relies on Active Bundle (AB), which incorporates policy enforcement mechanism and policies that describe access to video data. The usage of ABs ensures privacy of actors during disclosure of captured video in untrusted environment. Face recognition algorithm is used to identify sensitive data in videos and it is used in policies. We use four algorithms to process images that are captured by a vehicle's camera.","PeriodicalId":415692,"journal":{"name":"2015 IEEE 34th Symposium on Reliable Distributed Systems Workshop (SRDSW)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129341241","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Developing and operating a complex secure application with high assurance is difficult and requires experts. Security patterns and best practices have been proposed to assist architects in designing secure applications. However, these are usually written independently of the underlying platforms and operating environment. This leads to a gap between patterns and the platforms, and does not directly support the design-level analysis and verification of systems to be built on those platforms. We propose an approach to incrementally build an application design using design fragments, which are specializations of patterns for target platforms. Design fragments can be composed and reused during design, and directly support design-level security analyses and operation level concerns. We apply this approach in a case study of the design and analysis of a smart electricity meter. We show how the approach can be used to iteratively address threats.
{"title":"Building Secure Applications Using Pattern-Based Design Fragments","authors":"Paul Rimba, Liming Zhu, Xiwei Xu, Daniel W. Sun","doi":"10.1109/SRDSW.2015.12","DOIUrl":"https://doi.org/10.1109/SRDSW.2015.12","url":null,"abstract":"Developing and operating a complex secure application with high assurance is difficult and requires experts. Security patterns and best practices have been proposed to assist architects in designing secure applications. However, these are usually written independently of the underlying platforms and operating environment. This leads to a gap between patterns and the platforms, and does not directly support the design-level analysis and verification of systems to be built on those platforms. We propose an approach to incrementally build an application design using design fragments, which are specializations of patterns for target platforms. Design fragments can be composed and reused during design, and directly support design-level security analyses and operation level concerns. We apply this approach in a case study of the design and analysis of a smart electricity meter. We show how the approach can be used to iteratively address threats.","PeriodicalId":415692,"journal":{"name":"2015 IEEE 34th Symposium on Reliable Distributed Systems Workshop (SRDSW)","volume":"2008 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134634329","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
L. Lilien, L. B. Othmane, Pelin Angin, B. Bhargava, Raed M. Salih, A. DeCarlo
We propose application of Opportunistic Resource Utilization Networks (Oppnets), a novel type of Mobile Ad Hoc NETworks (MANETs), for ad hoc networking of Unmanned Aerial Vehicles (UAVs) in surveillance missions. Oppnets provide effective resource virtualization and adaption to highly dynamic and unstable nature of MANETs. They can be viewed as middleware to facilitate building flexible and adaptive distributed systems that provide all kinds of resources or services to the requesting application via the so called helper mechanism. The simulation study focuses on the impact of an initial target position on the performance of Oppnet-based UAV surveillance systems. We find that detection success ratios and time to detect a target are negligibly affected by the initial target position in the surveillance area when UAVs expand up their Oppnet quickly, but strongly affected by the initial target position when UAVs are slow in building up their Oppnet.
{"title":"Impact of Initial Target Position on Performance of UAV Surveillance Using Opportunistic Resource Utilization Networks","authors":"L. Lilien, L. B. Othmane, Pelin Angin, B. Bhargava, Raed M. Salih, A. DeCarlo","doi":"10.1109/SRDSW.2015.11","DOIUrl":"https://doi.org/10.1109/SRDSW.2015.11","url":null,"abstract":"We propose application of Opportunistic Resource Utilization Networks (Oppnets), a novel type of Mobile Ad Hoc NETworks (MANETs), for ad hoc networking of Unmanned Aerial Vehicles (UAVs) in surveillance missions. Oppnets provide effective resource virtualization and adaption to highly dynamic and unstable nature of MANETs. They can be viewed as middleware to facilitate building flexible and adaptive distributed systems that provide all kinds of resources or services to the requesting application via the so called helper mechanism. The simulation study focuses on the impact of an initial target position on the performance of Oppnet-based UAV surveillance systems. We find that detection success ratios and time to detect a target are negligibly affected by the initial target position in the surveillance area when UAVs expand up their Oppnet quickly, but strongly affected by the initial target position when UAVs are slow in building up their Oppnet.","PeriodicalId":415692,"journal":{"name":"2015 IEEE 34th Symposium on Reliable Distributed Systems Workshop (SRDSW)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117336923","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Mobile devices are subject to an increased attack surface vector as compared to desktop computing, due to the nature of sensors, radios, and increased peripherals. We investigate in this work mobile botnets with a specific focus on Android, which is the most widely adopted mobile platform, and a prime target for malicious software, 79% of reported malware threats to mobile operating systems are targeted at Android. Our analysis focuses on a short messaging service (SMS) botnet structure and investigates a new detection model using the concept of intents. We show that transparent control can be achieved by a remote endpoint, yet also detected by our proposed intent detection model. Intents are late run-time bindings mechanisms provided to applications in the Android operating system. Intents provide a clear and accurate picture of device behaviour with external sources, due to their design as a late run time binding mechanism in the Android Operating System. We propose an intent logging system to capture sample data, and use this as the basis to design and evaluate our proposed detection scheme.
{"title":"SMS Botnet Detection for Android Devices through Intent Capture and Modeling","authors":"Erik Johnson, I. Traoré","doi":"10.1109/SRDSW.2015.21","DOIUrl":"https://doi.org/10.1109/SRDSW.2015.21","url":null,"abstract":"Mobile devices are subject to an increased attack surface vector as compared to desktop computing, due to the nature of sensors, radios, and increased peripherals. We investigate in this work mobile botnets with a specific focus on Android, which is the most widely adopted mobile platform, and a prime target for malicious software, 79% of reported malware threats to mobile operating systems are targeted at Android. Our analysis focuses on a short messaging service (SMS) botnet structure and investigates a new detection model using the concept of intents. We show that transparent control can be achieved by a remote endpoint, yet also detected by our proposed intent detection model. Intents are late run-time bindings mechanisms provided to applications in the Android operating system. Intents provide a clear and accurate picture of device behaviour with external sources, due to their design as a late run time binding mechanism in the Android Operating System. We propose an intent logging system to capture sample data, and use this as the basis to design and evaluate our proposed detection scheme.","PeriodicalId":415692,"journal":{"name":"2015 IEEE 34th Symposium on Reliable Distributed Systems Workshop (SRDSW)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129828332","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Strongly consistent systems supporting distributed transactions can be prone to high latency and do not tolerate partitions. The present trend of using weaker forms of consistency, to achieve high availability, poses notable challenges in writing applications due to the lack of linearizability, e.g., to ensure global invariants, or perform mutator operations on a distributed datatype. This paper addresses a specific problem: the exactly-once transfer of a "quantity" from one node to another on an unreliable network (coping with message duplication, loss, or reordering) and without any form of global synchronization. This allows preserving a global property (the sum of quantities remains unchanged) without requiring global linearizability and only through using pairwise interactions between nodes, therefore allowing partitions in the system. We present the novel quantity-transfer algorithm while focusing on a specific use-case: a redistribution protocol to keep the quantities in a set of nodes balanced, in particular, averaging a shared real number across nodes. Since this is a work in progress, we briefly discuss the correctness of the protocol, and we leave potential extensions and empirical evaluations for future work.
{"title":"Exactly-Once Quantity Transfer","authors":"Ali Shoker, Paulo Sérgio Almeida, Carlos Baquero","doi":"10.1109/SRDSW.2015.10","DOIUrl":"https://doi.org/10.1109/SRDSW.2015.10","url":null,"abstract":"Strongly consistent systems supporting distributed transactions can be prone to high latency and do not tolerate partitions. The present trend of using weaker forms of consistency, to achieve high availability, poses notable challenges in writing applications due to the lack of linearizability, e.g., to ensure global invariants, or perform mutator operations on a distributed datatype. This paper addresses a specific problem: the exactly-once transfer of a \"quantity\" from one node to another on an unreliable network (coping with message duplication, loss, or reordering) and without any form of global synchronization. This allows preserving a global property (the sum of quantities remains unchanged) without requiring global linearizability and only through using pairwise interactions between nodes, therefore allowing partitions in the system. We present the novel quantity-transfer algorithm while focusing on a specific use-case: a redistribution protocol to keep the quantities in a set of nodes balanced, in particular, averaging a shared real number across nodes. Since this is a work in progress, we briefly discuss the correctness of the protocol, and we leave potential extensions and empirical evaluations for future work.","PeriodicalId":415692,"journal":{"name":"2015 IEEE 34th Symposium on Reliable Distributed Systems Workshop (SRDSW)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132542361","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In recent years, data replication have received attention in terms of tolerance for large-scale disasters. Also, file versioning features, by which more than one version of a file can be maintained, have been commercially available in some storage services. For providing file versioning with limited storage resources, it is essential to divide the resources among versions in accordance with the varied needs of numerous users. In this paper, we focus on applications in which newer versions of a file are more likely to be requested, which may be true in the case of many subscription services. We propose a new distributed data replication protocol supporting the file versioning feature. In order to eliminate a single point of failure, replica nodes themselves do not know about which nodes hold the newest version, instead, clients dynamically search the nodes having the required version in read/write operations. We also construct an analytical model that can derive an optimal allocation of the resources when the total number of replica nodes in a system and the distribution of the frequency of read requests for each version are given. In addition, we present some numerical examples obtained by simulations to show the good scalability and dependability of our system by assuming some realistic parameters.
{"title":"A Client-Based Replication Protocol for Multiversion Cloud File Storage","authors":"Mamoru Ohara, S. Fukumoto","doi":"10.1109/SRDSW.2015.17","DOIUrl":"https://doi.org/10.1109/SRDSW.2015.17","url":null,"abstract":"In recent years, data replication have received attention in terms of tolerance for large-scale disasters. Also, file versioning features, by which more than one version of a file can be maintained, have been commercially available in some storage services. For providing file versioning with limited storage resources, it is essential to divide the resources among versions in accordance with the varied needs of numerous users. In this paper, we focus on applications in which newer versions of a file are more likely to be requested, which may be true in the case of many subscription services. We propose a new distributed data replication protocol supporting the file versioning feature. In order to eliminate a single point of failure, replica nodes themselves do not know about which nodes hold the newest version, instead, clients dynamically search the nodes having the required version in read/write operations. We also construct an analytical model that can derive an optimal allocation of the resources when the total number of replica nodes in a system and the distribution of the frequency of read requests for each version are given. In addition, we present some numerical examples obtained by simulations to show the good scalability and dependability of our system by assuming some realistic parameters.","PeriodicalId":415692,"journal":{"name":"2015 IEEE 34th Symposium on Reliable Distributed Systems Workshop (SRDSW)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122002557","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We design a general framework named Hdoctor for hard drive failure prediction. Hdoctor leverages the power of big data to achieve a significant improvement comparing to all previous researches that used sophisticated machine learning algorithms. Hdoctor exhibits a series of engineering innovations: (1) constructing time dependent features to characterize the Self-Monitoring, Analysis and Reporting Technology (SMART) value transitions during disk failures, (2) combining features to enable the model to learn the correlation among different SMART attributes, (3) regarding circumstance data such as cluster workload, temperature, humidity, location as related features. Meanwhile, Hdoctor collects/labels samples and updates model automatically, and works well for all kinds of disk failure prediction in our intelligent data center. In this work, we use Hdoctor to collect 74,477,717 training records from our clusters involving 220,022 disks. By training a simple and scalable model, our system achieves a detection rate of 97.82%, with a false alarm rate (FAR) of 0.3%, which hugely outperforms all previous algorithms. In addition, Hdoctor is an excellent indicator for how to predict different hardware failures efficiently under various circumstances.
设计了一个用于硬盘故障预测的通用框架Hdoctor。与之前所有使用复杂机器学习算法的研究相比，Hdoctor利用大数据的力量实现了显著的改进。Hdoctor展示了一系列工程创新:(1)构建时间相关特征来表征磁盘故障时SMART (Self-Monitoring, Analysis and Reporting Technology)的值转换;(2)组合特征使模型能够学习不同SMART属性之间的相关性;(3)将集群工作负载、温度、湿度、位置等环境数据作为相关特征。同时，Hdoctor可以自动采集/标记样本并更新模型，可以很好地用于我们智能数据中心的各种磁盘故障预测。在这项工作中，我们使用Hdoctor从涉及220,022个磁盘的集群中收集了74,477,717条训练记录。通过训练一个简单且可扩展的模型，我们的系统实现了97.82%的检测率，虚警率(FAR)为0.3%，大大优于以前的所有算法。此外，对于如何在各种情况下有效地预测不同的硬件故障，Hdoctor是一个很好的指示器。
{"title":"Hard Drive Failure Prediction Using Big Data","authors":"Wenjun Yang, Dianming Hu, Yuliang Liu, Shuhao Wang, Tianming Jiang","doi":"10.1109/SRDSW.2015.15","DOIUrl":"https://doi.org/10.1109/SRDSW.2015.15","url":null,"abstract":"We design a general framework named Hdoctor for hard drive failure prediction. Hdoctor leverages the power of big data to achieve a significant improvement comparing to all previous researches that used sophisticated machine learning algorithms. Hdoctor exhibits a series of engineering innovations: (1) constructing time dependent features to characterize the Self-Monitoring, Analysis and Reporting Technology (SMART) value transitions during disk failures, (2) combining features to enable the model to learn the correlation among different SMART attributes, (3) regarding circumstance data such as cluster workload, temperature, humidity, location as related features. Meanwhile, Hdoctor collects/labels samples and updates model automatically, and works well for all kinds of disk failure prediction in our intelligent data center. In this work, we use Hdoctor to collect 74,477,717 training records from our clusters involving 220,022 disks. By training a simple and scalable model, our system achieves a detection rate of 97.82%, with a false alarm rate (FAR) of 0.3%, which hugely outperforms all previous algorithms. In addition, Hdoctor is an excellent indicator for how to predict different hardware failures efficiently under various circumstances.","PeriodicalId":415692,"journal":{"name":"2015 IEEE 34th Symposium on Reliable Distributed Systems Workshop (SRDSW)","volume":"11 4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116114743","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We present a new programming model for large-scale mobile and "Internet of Things" style distributed applications. The model consists of two layers: a language layer based on the Lasp language with a runtime layer based on epidemic broadcast. The Lasp layer provides deterministic coordination-free computation primitives based on conflict-free replicated data types (CRDTs). The epidemic broadcast layer is based on the Plumtree protocol. It provides a communication framework where clients may only have a partial view of membership and may not want to participate in or have knowledge of all active computations. We motivate the new model with a nontrivial mobile application, a distributed ad counter, and we give the model's formal semantics.
{"title":"Selective Hearing: An Approach to Distributed, Eventually Consistent Edge Computation","authors":"Christopher S. Meiklejohn, P. V. Roy","doi":"10.1109/SRDSW.2015.9","DOIUrl":"https://doi.org/10.1109/SRDSW.2015.9","url":null,"abstract":"We present a new programming model for large-scale mobile and \"Internet of Things\" style distributed applications. The model consists of two layers: a language layer based on the Lasp language with a runtime layer based on epidemic broadcast. The Lasp layer provides deterministic coordination-free computation primitives based on conflict-free replicated data types (CRDTs). The epidemic broadcast layer is based on the Plumtree protocol. It provides a communication framework where clients may only have a partial view of membership and may not want to participate in or have knowledge of all active computations. We motivate the new model with a nontrivial mobile application, a distributed ad counter, and we give the model's formal semantics.","PeriodicalId":415692,"journal":{"name":"2015 IEEE 34th Symposium on Reliable Distributed Systems Workshop (SRDSW)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116278579","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：